A. Introduction: “Exceptionalized Pluralism” as the Defining Feature of the 2024–2026 Reform of the EU Return Policy
The reform of the EU return acquis in 2024–2026 marks the most far–reaching transformation of the European return law since the adoption of Directive 2008/115/EC.Footnote 1 The legislative package adopted under the Pact on Migration and Asylum—most notably the Asylum Procedures Regulation,Footnote 2 the Return Border Procedure Regulation,Footnote 3 and amendments to the Schengen Borders CodeFootnote 4 —was followed in March 2025 by the Commission’s Proposal for a Regulation on common return procedures (PRR),Footnote 5 the Council’s agreed position of December 2025,Footnote 6 and the European Parliament’s plenary adoption of the consolidated text on June 17, 2026.Footnote 7 Together, these instruments seek to replace a minimum–harmonization model based on a directive with a directly applicable regulation premised on maximum procedural harmonization. This Article critically assesses the impact of that reform on EU return law as judicially constructed through almost two decades of the Court of Justice of the EU (CJEU) jurisprudence.
The adopted Regulation enters into force the day following its publication in the Official Journal of the European Union, but it does not apply as a whole from that date. Article 50 establishes a staggered scheme: the provisions that supersede Directive 2008/115/EC apply only twelve months after entry into force, while Article 50(2) carves out a closed list of provisions applicable from entry into force itself. That list is analytically revealing. It comprises the definitional and framing core—Article 1 on subject matter, Article 2(1) to (3) on scope, the definition of “country of return” in Article 3, point (3), and Article 4 on fundamental rights; the third-country agreement architecture that underpins the so-called return hub model, namely Articles 14, 15, and 16; the external dimension and readmission track in Articles 34 and 35; and a cluster of institutional and preparatory provisions, namely Articles 41, 43, 47, 48, and 49, together with Article 5(13) and (15). Recital 65 presents the deferral as a means of allowing the necessary preparatory and operational steps to be taken and of avoiding overlap with Directive 2008/115/EC. The sequencing is nonetheless substantive rather than merely technical. The instruments of externalization and diplomatic leverage are applicable a full year before the procedural provisions governing the return procedure itself, and the minimum-treatment floor in Article 2(7), which protects persons excluded from the Regulation’s scope under Article 2(4), points (a) and (b), is itself among the provisions deferred to month twelve.
The reform unfolds against a broader normative shift in EU migration governance. Scholarly accounts have characterized this shift as “concerted dis-integration,”Footnote 8 “crisification,”Footnote 9 “organised hypocrisy,”Footnote 10 “de–legalisation of policy,”Footnote 11 “normalisation of disintegration,”Footnote 12 and “securitisation and exceptionalism.”Footnote 13 While the securitization of migration law, discourse, and practice is a long-standing phenomenon, it has gained renewed momentum with the “crisification” and disintegration model of EU legislating in the field of asylum. It manifests, for instance, in the Commission’s justification of exceptions to the principle of non–refoulement—a non-derogable cornerstone of international lawFootnote 14 —in Member States’ legislative codification of emergency regimes suspending the right to asylum at external borders,Footnote 15 and in the overt disregard of court decisions affirming the primacy of EU law.Footnote 16
This Article argues that the 2024–2026 EU return reform, while exhibiting some traits of all the above-mentioned theoretical frameworks, does not fit squarely within any of them. The Article proposes “exceptionalized pluralism” as an overarching analytical framework: a structural logic through which the multiplication of derogatory procedural tracks—at external and internal EU borders and extraterritorially—collectively displaces the common return procedure and its Charter-based safeguards, transforming what was designed as exception into the operative norm of return governance. This logic manifests across two interconnected dimensions.
The first is an intensified proceduralization characterized by a proliferation of derogations from the common return procedure, progressively normalized to the detriment of the fall-back procedure and thereby eroding the procedural safeguards attaching to returnees’ rights—notwithstanding certain incremental improvements introduced under the reformed return procedure.Footnote 17
The second is a systematic recourse to externalization and coercive detention as the preferred instruments of return management, which together project exceptionalized pluralism extraterritorially and recalibrate the constitutional relationship between liberty and executive power beyond what Articles 6 and 19(2) CFR and the CJEU’s settled jurisprudence permit. Crucially, the constitutional risk does not lie in any individual derogation considered in isolation, although each of them raises specific challenges; it lies in their cumulative operation, which undermines the Charter-based standards—that two decades of judicial constitutionalization have rendered constitutionally non-negotiable—and hinders the leading objective of the reform, namely increased harmonization.
The return regulation proposals put forward by the Commission, the Council, and the Parliament have attracted heavy criticism from international organizations and civil society for failing to uphold fundamental rights and the EU’s rule of law threshold.Footnote 18 This is not without historical precedent; the 2008 Return Directive was similarly denounced at the time of its adoption,Footnote 19 yet over the following two decades the CJEU and national courts transformed it into a rights-sensitive framework through more than fifty preliminary rulings.Footnote 20 Through that process of judicial constitutionalization, the Court established limits on the criminalization of mere irregular stay, the principle of proportionality when using coercion in return procedures, the right to be heard, effective judicial remedies with suspensive effect, and the autonomous scope of non-refoulement under Article 19(2) CFR.Footnote 21 The 2024–2026 reform both inherits this jurisprudential baseline and is constitutionally bound to respect it.
This Article argues that the reform fundamentally reconfigures the EU return acquis in a manner that belies its promise of harmonization. Rather than producing uniform rights and procedures, the regulatory framework multiplies derogatory tracks and compresses substantive and procedural safeguards through extended grounds and temporal detention, accelerated border procedures, mutual recognition of return decisions without adequate human rights safeguards, and externalization of return enforcement through “return hubs.” Harmonization thus operates not as the uniformization of rights, but as the coordination of coercion. The central question is whether this exceptionalized pluralism model remains compatible with the Union’s foundational commitments to the effectiveness of EU law, rule of law, effective judicial protection, liberty, and the absolute prohibition of refoulement.
The argument proceeds along four interconnected lines that structure the Article. First, the shift from a Directive to a Regulation does not automatically produce genuine uniformity: the framework creates derogatory regimes—border procedures, fast-track intra-EU transfers, and expanded readmission agreements—that preserve significant Member State discretion while compressing procedural safeguards, such that harmonization risks entrenching uneven protection rather than eliminating it. Second, the European Return Order reproduces structural tensions familiar from the Dublin system, redistributing enforcement authority without fully clarifying the independent fundamental rights obligations of the executing State, and creating accountability gaps that the Court’s existing jurisprudence—in particular N.S. Footnote 22 and C.K. Footnote 23 —renders constitutionally untenable. Third, the expansion of pre-removal detention powers, and fourth, the formalization of return externalization through “return hubs,” signal a normative shift from proportionate and individualized return measures toward coercive enforcement, challenging the jurisprudential baseline according to which detention is exceptional, return should consider the connection to the receiving country, and the right to non-refoulement is absolute.
The Article introduces the concept of exceptionalized pluralism to capture a structural feature of the reform that existing scholarly frameworks have not fully theorized in the return-specific context: the codification of multiple legally authorized derogatory tracks that collectively displace, rather than supplement, the common return procedure and its Charter-based safeguards. The term is chosen deliberately. “Exceptionalized” marks the inversion by which authorized derogations become the operative norm rather than remaining exceptions from it, and the “pluralism” at issue is not the normative diversity familiar from accounts of constitutional pluralism but the multiplication of derogatory enforcement tracks. The Article argues that this exceptionalized pluralism has a de–constitutionalizing effect, namely each individual derogation may appear proportionate in isolation, but their cumulative operation undermines the constitutional baseline the CJEU has established as non-negotiable. Moreover, it shows that the externalization of returns without adequate constitutional safeguards amplifies judicial contestation rather than reducing it.
The Article proceeds as follows. Section B analyzes the contribution of European judicial interaction to the normative and human rights development of the common return procedure. Section C examines the structural transformations introduced by the 2024 Common European Asylum System (CEAS) reform through the lens of exceptionalized pluralism, focusing on the merger of asylum and return procedures and the differentiated border procedures at internal and external EU borders, that together displace the common return procedure safeguards. Section D critically assesses a novel mechanism under the new Return Regulation—the mutual recognition mechanism and the European Return Order—as the inter-state dimension of exceptionalized pluralism. Section E examines the expansion of pre-removal detention in the new Return Regulation as the coercive infrastructure that makes the derogatory tracks operationally viable. Section F analyzes the formalization of return externalization through the “return hub” model, using the Italy-Albania Protocol as a constitutional stress-test. The Conclusion evaluates whether the reconfigured return acquis consolidates the Union’s constitutional commitments or institutionalizes exceptionalized pluralism as a permanent structural feature of return governance. In either case, the reform leaves courts to act as gatekeepers of the rule of law and fundamental rights, a role they must increasingly discharge under acute political pressure.
B. The Judicial Constitutionalization of Return Governance
The development of EU return acquis over almost two decades reveals a structural transformation in the governance of irregular migration. What began as a minimum-harmonization instrument designed to facilitate removalFootnote 24 has, through sustained judicial engagement, evolved into a site of constitutional contestation. This transformation was not the product of comprehensive legislative redesign. Rather, it emerged incrementally through judicial interactions between national courts, the CJEU, and the European Court of Human Rights (ECtHR),Footnote 25 which progressively embedded return governance within constitutional commitments, such as proportionality, effective judicial protection, non-refoulement, the best interests of the child, the right to liberty, and the rule of law. The European judicial dialogue thereby recalibrated the relationship between executive discretion and fundamental rights in a domain historically under the Member States’ prerogative.Footnote 26
The 2008 Return Directive was adopted as a minimum-harmonization measure and deliberately preserved broad or vague language and extensive national discretion regarding the grounds for detention,Footnote 27 the design of remedies, and the interaction between administrative and criminal enforcement. These legislative choices generated structural fragility. Core procedural guarantees, such as the right to be heard, were absent altogether,Footnote 28 safeguards were general rather than operational,Footnote 29 such as the right to effective remedy instead of a right to effective judicial remedy,Footnote 30 and key questions—most notably the relationship between criminal sanctions and return procedures—were left unresolved.Footnote 31 This normative vagueness produced divergent national practices and uneven levels of protection across Member States.Footnote 32 At the same time, it created the conditions for judicial intervention: in the absence of legislative precision, courts were called upon to articulate the Directive’s internal logic and its relationship to EU primary law and fundamental rights instruments. Through the preliminary reference procedure under Article 267 of the Treaty on the Functioning of the European Union (TFEU), national courts and the CJEU became the principal architects of the return acquis, elaborating standards that the legislature had declined to specify and ensuring that Member States grounded their implementing measures in national legislation rather than executive orders inconsistent with the Directive’s framework.Footnote 33
The first domain of judicial constitutionalization concerned the relationship between criminal law and administrative return. Several Member States continued to impose custodial penalties for irregular stay or non-compliance with return orders, invoking their retained criminal competence.Footnote 34 In a line of judgments beginning with El Dridi Footnote 35 and followed by Achughbabian,Footnote 36 Sagor,Footnote 37 Celaj,Footnote 38 and Zh. and O.,Footnote 39 the CJEU established that national criminal sanctions must not undermine the effectiveness of the Directive. The Court held that imprisonment for mere irregular stay, where return procedures are available or ongoing, is incompatible with the Directive’s objective of establishing a coherent and incremental removal system. Criminal law may be used only after the exhaustion of the Directive’s procedural sequence, starting from the least intrusive measure, voluntary departure, and progressing to enforcement of removal and, where necessary, detention. Moreover, criminal law cannot function as a parallel or substitute regime.Footnote 40 This jurisprudence constitutionalized a structural hierarchy between administrative and criminal enforcement, grounding its reasoning in the EU law principles of primacy, effectiveness, sincere cooperation, and proportionality. The constitutional significance lies not in the abolition of criminalization as such, which remains permissible in defined circumstances, but in the imposition of principled constraints on its deployment.
A second domain of constitutionalization concerns pre-removal detention. Historically characterized by broad administrative discretion and deferential judicial review,Footnote 41 detention was restructured through judicial interpretation of Article 15 of the Return Directive. In El Dridi,Footnote 42 the Court emphasized that return enforcement must proceed with the least coercive measure available, establishing voluntary departure as the primary modality. Subsequent judgments—including Arslan Footnote 43 and Mahdi Footnote 44 —articulated more concrete standards, holding that detention is permissible only where it is necessary to prepare removal, where less coercive alternatives are insufficient,Footnote 45 and where there is a genuine risk of absconding or obstruction. The grounds for detention must, moreover, be laid down in national legislation that meets standards of transparency and legal certainty, precluding reliance on executive orders or administrative practice alone.Footnote 46 Detention must be proportionate, of limited duration, and pursued with diligence towards the objective of removal.Footnote 47 Most significantly, in Mahdi, FMS,Footnote 48 and X,Footnote 49 the CJEU expanded the scope of judicial review on the basis of Articles 6 and 47 CFR. National courts must be able to examine all relevant questions of fact and law, assess proportionality and necessity, and substitute their own decision for that of the administrative authority, including ordering release, even where national procedural law would not otherwise permit such a remedy. Through this line of authority, pre-removal detention—once a fall-back measure within return proceduresFootnote 50 —was reframed as an exceptional measure of last resort subject to EU constitutional constraints, with proportionality requiring strictly limited duration and the rule of law requiring clear legislative grounding and immediate release as a remedy.
A third domain concerns procedural safeguards more broadly. The Directive’s text contained limited express guaranteesFootnote 51 and, notably, omitted any specific provision on the right to be heard.Footnote 52 Through judicial interaction, however, such safeguards were progressively read into the return framework.Footnote 53 In Abdida,Footnote 54 Gnandi,Footnote 55 and CPAS cases,Footnote 56 the CJEU recognized an automatic suspensive effect of appeal where removal may expose an individual to a serious risk of treatment contrary to Articles 19(2) and 47 CFR, extending beyond what Article 13 of the Directive expressly stipulates. The Court further held in FMS that Article 13 must be interpreted as requiring judicial review before a court of law, not merely an administrative authority, notwithstanding the text’s apparent permissiveness on this point.Footnote 57 In Abdida, the Court extended protection in medical cases, holding that removal is prohibited where it risks serious and irreversible deterioration in health.Footnote 58 This reasoning subsequently influenced the ECtHR’s own recalibration of its jurisprudence in Paposhvili v. Belgium,Footnote 59 illustrating the transnational dimension of judicial constitutionalization.
A fourth domain concerns the protection of vulnerable persons, and in particular unaccompanied minors. In TQ,Footnote 60 the CJEU addressed a legislative lacuna between the issuance of a return decision and the act of removal. The Court held that Member States must assess the best interests of the child at all stages of the procedure and verify the availability of adequate reception in the country of return before a return decision may lawfully be issued.Footnote 61 This interpretation transforms Article 5 of the Directive from a programmatic statement into a substantive operational obligation, drawing upon Article 24 CFR and international child rights instruments.Footnote 62 Similarly, in X and Adrar,Footnote 63 the Court constitutionalized the principle of non-refoulement as an autonomous and structural constraint within EU return law, operative independently, and applicable at each stage of the return procedure. These judgments collectively constitutionalize vulnerable persons’ rights as substantive and procedural benchmarks binding on Member States throughout the return process.
Taken together, this jurisprudence operates on two interconnected levels. First, it transforms return proceedings into rights-sensitive procedures subject to due process guarantees derived directly from EU primary law, rather than from legislative specification alone. Second, it embeds non–refoulement as a transversal and structural constraint internal to EU return law, operative independently of, and in some respects beyond, the obligations imposed by the European Convention on Human Rights (ECHR) and the 1951 Refugee Convention. It is worth noting that this process of judicial constitutionalization, while significant in its scope, built upon foundations that the Directive itself had partially established. In several Member States, the national frameworks that the Directive replaced had criminalized mere irregular entry,Footnote 64 imposed disproportionate or lifetime entry bans,Footnote 65 or characterized pre-removal detention as a mere limitation of freedom of movement rather than a deprivation of liberty.Footnote 66 Measured against those regimes, the Directive marked a substantive improvement in important respects for the protection of returnees’ rights.
The more recent trajectory of constitutionalization has unfolded against a backdrop of intensifying political contestation. Post-2015 political discourse increasingly framed return policy in terms of enforcement deficits and inadequate return rates,Footnote 67 a framing that rested on contested statistical foundationsFootnote 68 and frequently elided the distinction between return orders issued and removals actually enforced. Member States openly criticized judicial interpretation of non-refoulement as expansive and as an obstacle to effective return.Footnote 69 It is within this political context that the legislative initiatives culminating in the 2024–2026 reforms must be assessed. Those reforms operate through an expanded pre-removal detention regime, extended derogations from common return procedures at internal and external borders, and extraterritorially via the formalization of return externalization. They directly challenge the jurisprudential baseline that two decades of judicial constitutionalization have established as constitutionally non-negotiable.
C. Exceptionalized Pluralism: The Impact of the 2024–2026 Return Reform
A central political promise of the return reform agenda has been that greater harmonization of return rules would enhance uniform national implementation, eliminate divergent practices, and replace a “patchwork” of national return systems with a genuinely common procedure.Footnote 70 Yet the legal architecture emerging from the 2024 instruments—most notably the Return Border Procedure Regulation and the revised Schengen Borders Code—and the 2026 Return Regulation point in the opposite direction. Rather than consolidating a single return procedure, these reforms institutionalize what this Article terms “exceptionalized pluralism”: a tiered system of derogatory return tracks in which the applicable safeguards vary according to three factors: geography, that is whether the person is apprehended at the border or inland; the procedural pathway, that is whether the post-border asylum procedure or the ordinary return procedure applies; and the logic of inter–state cooperation, that is whether readmission or transfer mechanisms are engaged.
The constitutional significance of this stratification lies not in differentiation as such—procedural flexibility to accommodate distinct factual circumstances is a recognized and legitimate legislative techniqueFootnote 71 —but in the direction and cumulative effect of the introduced derogations. Each derogatory track does not merely modify the common procedure; it systematically displaces the procedural safeguards that the CJEU has constitutionalized over close to two decades of judicial interaction with the Return Directive. The common return procedure does not disappear, but is progressively relegated to residual status, applicable only where no derogatory pathway is available. What was designed as the exception becomes the operative norm; what was designed as the norm becomes the fall-back. This structural inversion carries constitutional weight because the safeguards embedded in the common procedure—proportionate and individualized detention, effective judicial remedies with suspensive effect, and autonomous non-refoulement assessment guaranteed by automatic suspensive effect of appeal—are not optional features of procedural design. They are Charter-based requirements that apply regardless of the procedural track through which removal is pursued. Exceptionalized pluralism does not formally eliminate those requirements; rather, it creates systemic conditions in which they are structurally more difficult to access and enforce in practice.
This section critically analyzes three return acceleration measures introduced by the 2024–2026 reform that risk normalizing pushbacks at EU internal borders and entrenching the return border procedure as the operative default, thereby transforming the common return procedure from the standard framework into a residual exception.
I. Merging Asylum and Return Procedures: The Risk of Subsuming Non–Refoulement Within Asylum Reasoning
One of the most consequential innovations of the reform architecture is the operational merger of asylum determination and return enforcement. Under the Asylum Procedures Regulation, once an application for international protection is rejected as inadmissible, unfounded, or withdrawn, the competent authority must proceed, without delay, with the adoption of a return decision within the same procedure.Footnote 72 This institutional continuity is intended to address a longstanding implementation concern: the fragmentation between asylum and return systems, which frequently produced administrative gaps between rejection of a protection claim and initiation of removal proceedings.Footnote 73 The reform may offer genuine efficiency gains, because earlier adoption of return decisions can reduce periods of irregular stay and facilitate voluntary departure. However, procedural acceleration does not occur only through shorter deadlines, it results from the structural compression of decision-making stages that previously performed distinct functions: international protection determination and return enforcement. By linking rejection directly to removal, the reform risks transforming return from an independently assessed administrative act into the automatic consequence of asylum refusal.
The central constitutional concern is not that persecution-based claims will be inadequately examined—these remain exhaustively governed by the Qualification RegulationFootnote 74 —but that the compression of procedural stages forecloses independent assessment of non-refoulement obligations falling outside the scope of international protection. Such obligations derive autonomous force from Article 19(2) CFR and Article 3 ECHR and are absolute in character; they admit neither proportionality balancing nor derogation.Footnote 75
The scope of those obligations extends materially beyond the conditions for refugee status or subsidiary protection. The ECtHR has consistently held that Article 3 ECHR prohibits removal in circumstances such as serious illness or absence of adequate medical care, even where no intentional harm attributable to state authorities can be established, and therefore irrespective of whether the conditions of Article 15(b) of the Qualification Regulation are satisfied.Footnote 76 The CJEU has aligned itself with this approach. In X, the Court held that a return decision may not be adopted where removal would expose a seriously ill person to a real risk of rapid, significant, and irreversible deterioration in health or to intense suffering owing to the absence of appropriate treatment in the country of return, notwithstanding the absence of any entitlement to international protection.Footnote 77 In Ararat, the Court further confirmed that competent authorities are required to assess non-refoulement ex officio at the return stage whenever relevant information emerges, regardless of whether an asylum application has been definitively rejected.Footnote 78 These obligations equally arise where international protection has been withdrawn or has ceased; a person who no longer qualifies as a refugee or beneficiary of subsidiary protection may nonetheless be protected against removal if return would expose her to a real risk of ill–treatment contrary to Article 3 ECHR.Footnote 79 Beyond the absolute prohibition of refoulement, removal may equally be constrained where it would disproportionately interfere with family or private life. Both Article 8 ECHR and Articles 7 and 24(2) CFR require that family life considerations, including the best interests of any child affected, are independently assessed at the return stage.Footnote 80
The merged procedure must therefore incorporate a structurally distinct, independently conducted assessment of Article 19(2) CFR obligations at the return stage, one that is not subsumed within, or rendered otiose by, the rejection of an asylum claim. When return decisions are issued within the same procedure as asylum rejection, there is a structural risk that non–refoulement analysis becomes absorbed into asylum reasoning rather than independently verified.Footnote 81 Administrative efficiency thereby risks producing procedural inertia: once protection is denied, the transition to return may appear institutionally automatic, even though non-refoulement obligations require a legally distinct assessment.
This risk is compounded by Article 26(3) of the Return Regulation, which limits the court’s non-refoulement scrutiny to risks emerging from material contained in the file as supplemented or clarified during adversarial proceedings. This approach conflicts with both CJEU and ECtHR standards. As the ECtHR established in M.S.S. v. Belgium and Greece,Footnote 82 and as the CJEU reaffirmed in X, Ararat, and Adrar, public authorities, including reviewing courts, must assess non-refoulement risks ex officio on the basis of all relevant and publicly available information, irrespective of whether the individual has expressly raised the issue or adduced specific evidence. Conditioning judicial scrutiny on the administrative record normalizes a form of passive review that is structurally incompatible with the absolute character of Article 19(2) CFR.
Acceleration pressures are further compounded by the Asylum Procedure Regulation’s (APR) appeal architecture. Although Article 68 APR guarantees suspensive effect for combined asylum-return appeals, Article 67(7)(a) APR permits appeal deadlines of between five and ten days in accelerated or detention contexts. Where applicants face trauma, language barriers, or restricted access to legal assistance, as is characteristic of border or detention settings, such timelines risk rendering the right to an effective judicial remedy under Article 47 CFR largely theoretical. The CJEU and ECtHR have consistently held that effectiveness requires not only suspensive effect but adequate time and facilities to prepare a meaningful challenge where Article 4 CFR, respectively Article 3 ECHR risks are at issue.Footnote 83
The constitutionalization of EU return law through CJEU jurisprudence has established that return decisions cannot be treated as administrative sequels to asylum rejection; they must remain subject to autonomous scrutiny capable of preventing irreversible harm. The merger of the asylum and return procedure can have efficiency gains only if non-refoulement and considerations of family life and the child’s best interests are independently assessed.
II. Exceptionalized Pluralism at Internal Borders
Article 6(3) of the Return Directive permits Member States to refrain from issuing a return decision where a person is taken back by another Member State pursuant to a bilateral readmission agreement or arrangement existing at the date of the Directive’s entry into force. Originally framed as a transitional standstill clause, allowing pre–existing readmission frameworks to continue while the common procedure entered into force, Article 6(3) has in practice operated as an enduring governance technique.Footnote 84 Internal–border readmission arrangements became a convenient legal vehicle for rapid inter–state transfers, particularly during periods of heightened political pressure around secondary movement within the Schengen area. The difficulty is not that inter–state cooperation is illegitimate; it is that cooperation under Article 6(3) of the Return Directive has repeatedly functioned as a procedural shortcut that bypasses the common return procedure. The combination of speed, informality, and limited individualized assessment creates conditions conducive to summary return, including transfers executed without meaningful information provision, without assessment or registration of asylum claims, and without robust non–refoulement scrutiny.Footnote 85 That outcome is not an incidental implementation flaw; it is structurally encouraged by the derogation’s logic, which shifts procedural weight away from the Directive’s safeguards and toward border police cooperation practicesFootnote 86 that prioritize speedy intra–border transfers over rights.
EU and national courts have consistently found that this implementation of readmission arrangements violates both EU return acquis and the fundamental rights guarantees of non–refoulement and effective judicial protection.Footnote 87 The CJEU has emphasized that derogations from the common return framework must be interpreted strictly and cannot be deployed to neutralize the Directive’s protective logic; therefore, internal border controls will remain such even after the re–introduction of internal border controls.Footnote 88 The Return Regulation has not incorporated these constitutional requirements into a revised Article 6(3). On the contrary, it eliminates the temporal limitation on the conclusion of readmission agreements, thereby entrenching and expanding a derogation that courts have consistently identified as structurally deficient.Footnote 89
The normalization of intra–EU transfers is further consolidated by the revised Schengen Borders Code, which introduces Article 23a, a procedure enabling one Member State to transfer a third–country national apprehended near an internal border to the neighboring Member State from which that person is presumed to have arrived.Footnote 90 The provision presents itself as a “Schengen–friendly” alternative to formal border reintroduction—an administrative channel that preserves free movement while enabling rapid inter–state cooperation. From a rule–of–law perspective, however, Article 23a is best understood as the codification of an enforcement practice previously contested as unlawful. Before codification, many internal transfers operated through opaque bilateral police arrangements, often unpublished and executed at speed. Courts in several Member States scrutinized such practices, consistently identifying deficiencies in individual assessment, failures to provide access to procedures, and the risk of collective expulsion.Footnote 91
Article 23a provides for some basic procedural safeguards: written transfer decisions, time limits, a nominal right of appeal—but without suspensive effect of appeal—and categorical exclusions. That formalization represents an improvement over the sheer informality in Article 6(3) readmission agreements. It does not, however, strengthen the common return procedure. It creates a further parallel channel, an internal Schengen transfer regime with thinner safeguards than the Directive’s ordinary procedure and a structural bias toward immediacy. Article 23a’s short execution window, operationalized as twenty–four hours, is not merely an efficiency feature; it is a safeguard–reducing mechanism. When transfer is designed to occur almost immediately, practical access to counsel, interpreters, and information about asylum possibilities becomes structurally compromised.Footnote 92 A remedy that exists formally but cannot be meaningfully exercised within the available timeframe is precisely what European courts have repeatedly condemned as ineffective.Footnote 93 Where transfer risks exposing an individual to refoulement—directly or through chain transfer—the right to an effective remedy requires the availability of automatic suspensive effect.Footnote 94 Article 23a, if not coupled with such suspensive effect and a limit on successive transfers, risks enabling irreversible harm before meaningful review can occur.
The categorical exclusions in Article 23a—covering applicants for international protection and beneficiaries of protection—depend for their effectiveness on accurate identification at the point of apprehension. Internal–border encounters are precisely the context in which identification is least reliable due to obstacles including language barriers, absence of documentation, accelerated processing, and administrative refusal of registering asylum claims increase the probability that protection needs are missed.Footnote 95 A mechanism designed for speed in conditions of known screening weakness predictably reproduces the risks of unlawful summary removal that courts have criticized under both external–border and internal–transfer related case law.Footnote 96
This analysis illustrates the broader logic of exceptionalized pluralism. Article 23a does not eliminate divergent national practice; it introduces a new EU–level derogation that will itself be implemented divergently, depending on national cultures of police cooperation, judicial oversight, and institutional capacity. The foreseeable result is not harmonization but a new category of litigation concerning the Charter conformity of transfer decisions, the adequacy of non–refoulement assessments in fast–track contexts, and the lawfulness of transfers without automatic suspensive effect of appeal. In institutionalizing a derogatory track below the constitutional baseline established by the CJEU, the legislature once again legislates into foreseeable judicial scrutiny, replicating the cycle in which courts are required to reconstruct safeguards that the political process has designed too thinly. However, that cycle of foreseeable litigation is not merely an institutional failure; it is borne disproportionately by vulnerable categories of returnees—those without documentation, legal assistance, or linguistic access—who are least equipped to activate judicial protection and most exposed to irreversible harm before courts can intervene.
III. Mandatory Return Border Procedure: From Exceptional Border Track to Fall–Back Procedure
The Return Border Procedure Regulation (RBPR) represents another example of exceptionalized pluralism. Whereas the border return track was previously optional for Member States, the RBPR makes it mandatory in defined circumstances: wherever an application for international protection has been rejected in the course of an asylum border procedure, the ensuing return must be conducted under the border track rather than the ordinary procedure. At a late stage in the negotiation of the 2024 CEAS package, the return border procedure was detached from the Asylum Procedures Regulation and placed in a separate, self–standing instrument, owing to its implications for the Schengen acquis. It operationalizes a dual–track return system: individuals processed at the border following a rejection in an asylum border procedure are subject to an accelerated border track,Footnote 97 while third–country nationals apprehended inland remain governed by the Return Directive’s ordinary procedure. The border procedure is defined by three interlinked features: the legal fiction of non–entry, compressed procedural deadlines, and a structural predisposition toward detention or confinement. Each of these features has been contested in European migration jurisprudence precisely because each tends to convert exceptional measures into routine practice, evading the safeguards judicially embedded in the common return procedure.Footnote 98
The RBPR is not a supplementary mechanism. It redefines which procedure will in practice be most frequently applied. By placing return in seamless continuity with border asylum rejection, it shortens stages that previously created space for safeguards such as time to prepare an appeal, obtain legal advice, gather evidence relevant to non–refoulement risks, or pursue voluntary departure.Footnote 99 Under Article 4(5) RBPR, voluntary departure is only available upon the explicit request of the person concerned, for a maximum of fifteen days. The RBPR’s compressed timeframe and containment setting further marginalize this pathway, rendering it dependent on the individual’s own initiative rather than a default facilitated by the administration. As the CJEU confirmed in Al Hoceima, voluntary return is a structurally significant component of a rights–compliant return system.Footnote 100 Its marginalization is therefore a constitutional deficiency, undermining both the Directive’s gradual architecture and the principle that coercive measures must represent the least intrusive means capable of achieving legitimate aims. Even where judicial review is formally strengthened in the APR and the RBPR relative to the 2018 Commission recast proposal,Footnote 101 the capacity to exercise that review effectively is structurally undermined by the acceleration and containment features of the border model.
Under the Return Directive’s ordinary procedure, as reconstructed through CJEU jurisprudence, deprivation of liberty is treated as exceptional and subject to rigorous proportionality analysis.Footnote 102 The RBPR departs from this baseline. Article 5 permits detention authorized during asylum proceedings to continue automatically into the return phase without additional justification, while Article 5(3) introduces additional grounds—public order and national security—absent from Article 15 of the Return Directive.Footnote 103 Even where the instrument employs the language of last resort, confinement is the operational condition that makes the procedure function, given the nexus with detention under the asylum border procedure. This is a normative transformation: when detention is the condition of possibility for accelerated return, the legal presumption shifts—liberty becomes contingent, and coercion becomes the default of return governance rather than its exception.
The RBPR thus institutionalizes a governance hierarchy in which border return constitutes a high–coercion, high–speed track, whereas the common return procedure becomes operative only for those whose irregular status arises inland. Political incentives, particularly under a discourse of low return rates, favor progressive expansion of the categories of persons channeled into the faster, more coercive border track.Footnote 104 Over time, the exception risks becoming the fall–back norm. This is precisely the consolidation of the logic of exceptionalized pluralism—not the formal abolition of the common procedure, but the systematic conditioning of its safeguards, being routed into an exceptional regime.
D. Mutual Recognition of Return Decisions and the European Return Order: Blind Mutual Trust
A central innovation of the Return Regulation is the introduction of a mechanism for the mutual recognition and enforcement of return decisions across Member States (Article 7), operationalized through a standardized European Return Order (ERO). The two must be distinguished: in the adopted text the ERO is not itself the recognition mechanism but the information instrument that feeds it, a common form containing the essential elements of a return decision, made available through the Schengen Information System,Footnote 105 while mutual recognition is the separate enforcement mechanism established by Article 7.Footnote 106 The ERO is explicitly framed as an instrument to prevent secondary movements of irregularly staying third–country nationals—a rationale articulated, for the first time, as a structural objective of the EU return policy.Footnote 107 Yet the evidentiary foundations of this objective remain uncertain. Secondary movements of irregular migrants are difficult to quantify, not least because they are intertwined with informal readmission practices and internal border controls for which transparent and comparable data are scarce.Footnote 108 In the absence of a comprehensive impact assessment analyzing the scale and structural drivers of such movements, the introduction of a far–reaching mutual recognition mechanism appears normatively ambitious relative to its empirical basis.Footnote 109 The ERO thus raises a broader constitutional question that runs through this section: whether efficiency–driven harmonization can proceed without a parallel consolidation of procedural and fundamental rights guarantees, a question to which the existing jurisprudence of the Court of Justice provides a clear answer under the Dublin system jurisprudence.Footnote 110
Under the Commission’s proposal, once a Member State issues a return decision, it must generate a standardized ERO containing the essential elements of that decision and, where applicable, the removal order. The ERO would be uploaded into the Schengen Information System (SIS) pursuant to Regulation (EU) 2018/1860 or exchanged through a dedicated information channel, with an implementing act establishing the ERO template and Member States and EU agencies ensuring technical operability of automated recognition and enforcement. The agreed text abandons the Commission’s mandatory model altogether. Under Article 7(1) the enforcing Member State “may” recognize and enforce a return decision issued elsewhere; recognition need take no administrative form (Article 7(2)), and a State that declines simply issues its own decision (Article 7(3)), so that neither recognition nor its refusal constitutes a reviewable act and any appeal lies only in the issuing State. Mandatory recognition is deferred to a possible future Commission proposal following a three–year implementation review (Article 7(9)–(10)).Footnote 111 The grounds on which recognition may be withheld—conflict with public policy, higher national protection for unaccompanied minors or trafficking victims, a pending appeal in the issuing State, incomplete information, or a different envisaged destination—appear only in the non–binding recitals.Footnote 112 While this discretionary approach may be presented as institutional prudence, it effectively preserves a structural accountability gap: by leaving recognition optional rather than codifying binding refusal grounds, the legislature avoids specifying the conditions under which the enforcing State must refuse execution on fundamental rights grounds, precisely the clarification that constitutional coherence requires.Footnote 113
Furthermore, the Return Regulation provides that the six-month implementing act deadline pulls the preparatory obligations for mutual recognition forward of the twelve-month applicability of Article 7 itself. The infrastructure of mutual recognition must therefore be built before the provision mandating it is in force.
The ERO draws structural inspiration from the mutual recognition logic of the Dublin transfer system, as well as from earlier instruments including Council Directive 2001/40/EC on the mutual recognition of expulsion decisions.Footnote 114 Like the Dublin system, it relies on mutual trust, centralized information exchange via SIS, and the presumption of equivalent procedural and fundamental rights protection across Member States. However, the analogy is only partial, and the distinction is legally consequential. Under Dublin, mutual recognition determines responsibility for examining an asylum claim, resulting in a transfer to another Member State that remains within the Union’s legal order and subject to its oversight mechanisms. Under the ERO framework, by contrast, the enforcing Member State executes removal to a third country on the basis of a decision it did not itself adopt. The accountability configuration is therefore fundamentally altered: the enforcing State exercises public authority with irreversible extraterritorial consequences, while relying on an assessment conducted elsewhere, without necessarily conducting its own substantive reassessment of the individual’s situation. This is the rationale of mutual recognition; however, other mutual recognition–based fields do not directly affect the principle of non–refoulement as the return procedure.Footnote 115 This pattern of dysfunction is not without precedent. Earlier mutual recognition instruments in EU migration law, including Directive 2001/40/EC, encountered significant implementation obstacles that progressively undermined their operational coherence, a trajectory the Meijers Committee has identified as a structural risk equally applicable to the ERO.Footnote 116
This redistribution of decision–making authority without a corresponding redistribution of legal responsibility risks reproducing the structural tensions that characterized early Dublin jurisprudence—and that the Court of Justice has since resolved in ways that directly constrain the ERO’s operational logic. In N.S. and Others, Footnote 117 C.K., Footnote 118 and Jawo,Footnote 119 the Court established that mutual trust cannot operate as an irrebuttable presumption where there are substantial grounds for believing that removal would expose the individual to a real risk of treatment contrary to Article 4 CFR. In Ararat, the Court further held that the executing authority must conduct an ex officio reassessment of non–refoulement obligations prior to removal, independently of the issuing authority’s prior determination.Footnote 120 In X, it clarified that Article 19(2) CFR prohibits removal where there is a substantial risk of grave and irreversible deterioration in health, even outside the framework of subsidiary protection.Footnote 121 Taken together, these cases establish a coherent principle: fundamental rights obligations under the Charter attach to the authority executing removal, and not only to the authority that adopted the initial decision. Both issuing and enforcing States remain independently bound, and this independent responsibility cannot be displaced by the operational logic of mutual recognition when the authority knows or ought to have known about potential risks of violation of the principle of non–refoulement.
The Regulation does not adequately reflect this jurisprudential framework. Because the country of return needs to be designated only before removal (Article 10(4)), the non–refoulement assessment is displaced to the removal stage, where it must be pursued on a 14–day appeal (Article 27) that carries no automatic suspensive effect (Article 28). The Regulation neither explicitly codifies the circumstances under which the enforcing State must suspend or refuse execution on fundamental rights grounds, nor provides a mechanism enabling that State to adopt a new decision where the issuing State’s assessment is found to be procedurally or substantively deficient. This normative gap is further compounded by the Regulation’s broadened definition of “country of return.” Article 3, point (3) of the Return Regulation expands permissible destinations beyond the country of origin or transit to include any country where the person has a right of entry, a “safe third country,” a “first country of asylum,” or a State linked through readmission or transfer arrangements. These categories—particularly “safe third country” and “first country of asylum”—have generated extensive and unresolved litigation in both asylum and return contexts.Footnote 122 Under the ERO system, the enforcing State may be required to execute removal to a destination determined by the issuing State’s application of these contested categories, without any obligation of independent substantive review. The risk of refoulement is accordingly structural rather than incidental. The appropriate legislative model is the post–MSS Dublin framework as judicially reconstructed at both systemic and individual levels. The Regulation should accordingly codify a binding refusal clause obliging the enforcing State to suspend or decline execution of an ERO where substantial grounds exist for believing that removal, including to the designated country of return, would breach Articles 4 or 19(2) CFR at either the systemic or individual level, irrespective of the issuing State’s prior assessment.Footnote 123
The ERO constitutes the most far–reaching attempt to Europeanize return enforcement to date, and its operational promise is real. Reducing duplication of procedures and addressing the practical consequences of onward movement within the Schengen area are legitimate objectives, and SIS integration may enhance administrative coherence. However, harmonization through mutual recognition cannot displace the constitutional commitments that have emerged through a decade of judicial consolidation by the Court of Justice in the field of the Dublin system.
Return decisions are not merely administrative acts; they are exercises of public authority engaging core Charter rights, including human dignity, protection against refoulement, effective judicial protection, and respect for family life. The retention in the adopted text of purely discretionary recognition implicitly acknowledges these structural vulnerabilities but stops short of creating any binding ground for refusing or suspending execution on fundamental rights grounds.
To ensure coherence with existing jurisprudence, the Return Regulation should explicitly provide that the enforcing Member State must refuse or suspend execution of an ERO where substantial grounds exist for believing that removal would entail a violation, at a systemic or individual level, of Article 4 or Article 19(2) CFR. It should further clarify that the enforcing authority may, where necessary, issue a new return decision to remedy procedural or substantive deficiencies identified during execution. Without these clarifications, the ERO risks achieving administrative efficiency at the cost of legal accountability before the ECtHR. More broadly, the ERO is precisely the instrument where a codified exception to automatic mutual recognition, grounded in Articles 4 and 19(2) CFR, would have been the constitutionally coherent response.
E. Increased Pre–Removal Detention: From Last Resort to Operative Baseline
The transformation of pre–removal detention is the most constitutionally significant dimension of the Return Regulation. What is at stake is not a marginal expansion of administrative powers but a structural shift in the normative baseline of return governance. Detention is progressively reconfigured from a residual measure, permissible only under strict necessity and as a last resort, into a primary instrument through which return enforcement is operationalized across the border procedure and the common return procedure in the Regulation. Article 6 CFR and Article 5 ECHR do not merely require procedural regularity; instead, they demand that detention remain exceptional, individualized, proportionate, and strictly connected to a legitimate aim, with robust judicial control capable of preventing executive coercion from becoming routinized.
The constitutional baseline is established by the 2008 Return Directive, which conceived detention under Article 15 as a last resort, permissible only where less coercive measures cannot effectively be applied and for two circumscribed purposes: a risk of absconding, or conduct that avoids or hampers preparation of return.Footnote 124 This design reflects the Directive’s proportionality logic: deprivation of liberty is justified only insofar as strictly necessary to secure removal feasibility, and only for as long as removal remains genuinely possible within a reasonable time. The CJEU has constitutionalized this framework through two particularly salient lines of jurisprudence. In Kadzoev, the Court held that detention can only be justified for as long as removal arrangements are in progress and a reasonable prospect of removal exists, and that detention cannot be justified by general migration–control objectives, deterrence, or administrative convenience.Footnote 125 In Al Chodor and I.L., the Court linked the detention power to the rule–of–law requirements of clarity, predictability, and protection against arbitrariness. The criteria for assessing the risk of absconding must be defined in binding national law, and a general risk of compromised enforcement without specifically defined grounds fails this test.Footnote 126 Together, these cases establish a concrete constitutional baseline: detention must be grounded in precise legal criteria, necessary in the individual case, not presumed, proportionate and least intrusive, and temporally bounded by realistic prospects of removal.
The 2024 Return Border Procedure Regulation departs from this baseline by entrenching a containment–oriented model in which deprivation of liberty becomes structurally embedded in the functioning of accelerated border procedures. Its architecture relies on continuity of confinement from the asylum border procedure into the return phase, combined with compressed deadlines and the legal fiction of non–entry.Footnote 127 The practical effect is that detention is rendered less a carefully justified measure than the default setting enabling the border procedure to operate. Under the ordinary Return Directive model, detention is the exceptional response to a risk of hindering removal. Under the RBPR model, containment becomes the procedural precondition for speed, which exemplifies how exceptional tracks shift the normative baseline.
The Return Regulation consolidates and extends this shift across all return tracks, risking the transformation of pre–removal detention from a constitutionally bounded exception into a de facto rule. Detention, moreover, sits within an enlarged architecture of coercion comprising an extensive duty to cooperate (Article 18), backed by investigative powers (Article 21) and consequences for non–compliance (Article 22). Within that architecture, non–cooperation operates as the trigger routing an individual between alternatives to detention and custody, so that the measures reinforce one another rather than forming a hierarchy of ascending severity. Four specific developments illustrate this structural recalibration.
First, Article 29(3) broadens the grounds for detention beyond the Directive’s two grounds of absconding and obstruction to six: risk of absconding; avoiding or hampering readmission, preparation, or removal; security risks within the meaning of Article 13; verification of identity or nationality; non–compliance with the cooperation obligations of Article 18(3); and—most problematically—an open–ended category of “other relevant grounds related to return, provided for in national law,” qualified only by a requirement that they be “necessary and proportionate.” The security–risk ground is itself expansive, extending under Article 13(1) to any threat “identified under national law.” The adopted text further permits detention to be maintained until departure even where the individual is cooperating with voluntary return (Article 30(6)).Footnote 128 These expansions weaken the removal nexus and introduce preventive rationales potentially incompatible with Article 6 CFR.
Second, Article 19 reworks absconding through rebuttable presumptions, which arise, among others, from unauthorized entry or movement into a Member State (Article 19(1)(a)) or from a return decision issued by another Member State (Article 19(1)(b)). These presumptions are supplemented by an overall–assessment list that includes the lack of a fixed address and the “lack of social anchoring, in particular the absence of family ties, and insufficient means of subsistence” (Article 19(2) (a) and (f)). The result is an over–inclusive legal construct in which structural vulnerability becomes evidence of evasiveness. When such indicators are elevated into presumptions, detention is triggered not because of demonstrable individual conduct, but because of migrant precarity as such, a result that sits uneasily with the prohibition of arbitrariness underlying both Article 6 CFR and Article 5 ECHR.
Third, the extension of maximum detention from the Directive’s eighteen–month ceiling to twenty–four months (Article 30(3)), combined with further extension periods of up to six months after that ceiling where a reasonable prospect of removal re–emerges (Article 30(4)), as well as longer court–set periods for those detained as security risks (Article 13(5)(d)), represents normative escalation rather than merely quantitative change. More critically, the rule that a fresh detention period applies where a person has moved to another Member State (Article 30(3)) enables serial detention across jurisdictions. Detention may be legally reset even though the individual’s liberty deprivation continues in substance, directly contradicting the logic of Kadzoev.Footnote 129
Fourth, the Regulation codifies procedural safeguards: reasoned written decisions in an accessible language, periodic review of detention at least every six months, and every three months for unaccompanied minors and for extensions beyond twenty–four months, detention conditions echoing Bero and Bouzalmate and Landkreis Gifhorn, and a catalogue of alternatives.Footnote 130 Nevertheless, these safeguards cannot be assessed in isolation from the overall structural design increasing coercive measures. Judicial review risks becoming an exercise in legitimizing detention where grounds are overly broad and presumption–driven, unless relying on EU primary legal bases in order to provide genuine constraint under expansive categories such as security risks or open–ended national grounds. Equally, alternatives to detention risk functioning less as liberty–preserving substitutes than as an extension of control if, as Article 30(2) and (5) permit, they may be imposed even where the grounds for detention are no longer fulfilled.Footnote 131
Read as a whole, the 2024–2026 reform trajectory shifts EU return governance from a system in which detention is exceptional and tightly boundedFootnote 132 towards one in which detention becomes a routine instrument of governance across all procedural tracks. This is not an incidental side effect but an emergent structural logic: border procedures require containment to function; secondary movement is sanctioned through presumptive detention grounds; and the Regulation broadens substantive grounds, extends temporal scope, and introduces presumptions that make detention easier to justify. This is the detention–specific manifestation of exceptionalized pluralism: where the border procedure normalizes containment as a procedural precondition, and where the Return Regulation broadens and presumes detention grounds based on the precarity of irregular status. While Article 6 CFR and the CJEU require pre–removal detention to be a last resort, under the reform it becomes, through cumulative architectural choices, the operative baseline of return governance, and courts are once again left to reconstruct constitutional limits that legislative design has failed to embed.
F. Return Hubs and the Extraterritorial Dimension of Exceptionalized Pluralism
One of the most constitutionally consequential innovations of the Return Regulation is the formalization of return enforcement externalization through what has been described as “return hubs.”Footnote 133 Although the Regulation avoids this terminology,Footnote 134 the combined operation of Article 3, point (3)(g) and Article 14 Regulation establishes, for the first time, a comprehensive Union–law architecture for the offshore execution of return enforcement, enabling Member States to transfer individuals to third–country facilities pending onward removal on the basis of an agreement or arrangement, irrespective of any connection between the individual and the hosting state. This represents the extraterritorial dimension of exceptionalized pluralism: a regime designed to accelerate return enforcement through geographical displacement of coercive powers, which, as the Italy-Albania Protocol already demonstrates, generates challenges for EU primary law conformity, in particular the rights of defense, specific rights of vulnerable categories—for example, right to family life—effective judicial remedies, and independent monitoring.Footnote 135
Article 3, point (3)(g) of the Regulation marks a paradigmatic shift from the traditional connection–based approach to return destinations. Under the Return Directive, permissible destinations were anchored in the individual’s relationship to the destination state including country of origin, transit country pursuant to a readmission agreement, or a third country voluntarily chosen by the individual.Footnote 136 Article 3, point (3)(g) introduces a new category of “any other third country” willing to accept the person pursuant to an agreement or arrangement. The legal justification for removal thus no longer flows from the individual’s connection to the destination state, but from the existence of an intergovernmental instrument whether in the form of a treaty or informal arrangement.
This shift is furthermore consequential due to the interaction of the Regulation with the reformed asylum acquis: Article 3, point (3)(e) Regulation recognizes as a country of return a safe third country used to ground inadmissibility under the Asylum Procedures Regulation.Footnote 137 As the safe third country concept is progressively broadened, most notably through the relaxation of connection requirements, the range of permissible return destinations expands correspondingly, intensifying the burden placed on judicial oversight under Article 47 CFR precisely in contexts where effective remedies are most difficult to guarantee. The adopted text does, however, narrow one dimension of this risk: Article 14(3) confines onward return from a hub to the connection–based destinations in Article 3, points (3)(a)–(d), excluding “safe third country” and “first country of asylum” as permissible onward destinations, a limitation that constrains chain–refoulement from the hub but leaves untouched the breadth of the initial designation. This is exceptionalized pluralism operating extraterritorially. What was designed as an exceptional residual category—return to a state with no connection to the individual—is normalized as a primary enforcement instrument, while the connection–based safeguards of the Return Directive are relegated to residual status.
A constitutionally significant feature of Article 14 of the Regulation is the provision of a dual category of instruments—formal “agreements” and informal “arrangements”—for the establishment of “return hubs.”Footnote 138 EU agreements concluded within the scope of Union competence ordinarily trigger the procedural framework of Article 218 TFEU, including Council authorization, Parliamentary involvement, transparency obligations, and ex ante and ex post judicial review. Informal arrangements operate outside this constitutional framework, evading parliamentary scrutiny and judicial scrutiny. Furthermore, the proposal allows for return hubs to be concluded by both the EU and Member States, thus allowing informality also at the national level.
Furthermore, Article 50(2) of the Return Regulation anticipates the operation of this framework compared to other articles. The provision governing return to a third country with which there is an agreement or arrangement applies from the date of entry into force, together with the definition of “country of return” on which it depends and the child-specific safeguards on best interests and age assessment in Articles 15 and 16, whereas Article 5 on the return decision and Article 10 on removal apply only twelve months later. Member States may therefore negotiate and conclude the agreements and arrangements the provision contemplates—already bound by its substantive conditions, its notification duties, and its exclusion of unaccompanied minors—throughout a year in which no return can lawfully be effected under them. This temporal architecture is enabling rather than operational. Namely, it secures twelve months of treaty-making before the enforcement machinery is switched on, and it does so at a stage when the procedural provisions that would govern any resulting transfer are not yet applicable. The temporal design thus mirrors the substantive one, privileging the construction of externalized enforcement capacity.
That said, Article 14(1) Regulation contains an important, if underenforced, constitutional constraint. It conditions agreements and arrangements on the third country’s respect for international human rights standards and principles, now expressly including the principle of non–refoulement (Article 14(1)). Rigorously applied, this requirement would constitute a significant limit, namely excluding arrangements with states whose protection framework presents systematic deficiencies in international human rights protection and requiring ex ante verification that the destination country has not been found to be in violation of international human rights standards by international bodies and national courts. Unaccompanied minors are excluded from transfer to an Article 14 third country (Article 14(5)); families with children are not.
Recital 34 further clarifies that operating a return hub constitutes implementation of Union law within the meaning of Article 51(1) CFR, thereby rendering the Charter fully applicable to offshore enforcement.Footnote 139 This is constitutionally significant: it means that Charter rights—including Articles 4, 6, 7, 18, 19(2), 24, and 47 CFR—bind the Member State’s and EU’s authorities in the design, conclusion, and execution of hub–based returns, and cannot be contractually displaced to a third country.
Whether this safeguard will be effective in practice depends, however, on its institutional operationalization.Footnote 140 Notably, the independent monitoring body required by Article 14(3)(d), and then only where the arrangement provides for onward return, remains insufficiently specified as regards its powers, independence, and access. As the monitoring requirement under Article 12 of the Regulation illustrates, such elements are indispensable to render oversight mechanisms operationally effective.
Furthermore, the institutional operationalization is weakened still further by the Council’s position, which explicitly tolerates territorial or category–based deficiencies in the third country’s protection framework subject to abstract “sufficient guarantees,”Footnote 141 and the adopted Article 14(3) expressly permits the arrangement to lay down “conditions for detention in the third country,” entrenching offshore detention as an operational feature of hub arrangements. The Regulation moreover leaves unresolved the temporary and conditional nature of the stay in a hub, namely whether time limits apply, and what follows if onward removal cannot be carried out.Footnote 142
The legal challenges posed by the EU model of “return hubs” are not merely theoretical. The Italy-Albania Protocol, explicitly invoked by Commission President von der Leyen as an inspiration for the Regulation’s hub architecture, provides a direct and instructive stress–test.Footnote 143 Signed in November 2023 and ratified in February 2024,Footnote 144 the Protocol established extraterritorial asylum processing centers under Italian jurisdiction, applying the accelerated border procedure to asylum seekers from designated safe countries of origin. Its implementation was halted from the outset. Since October 2024, Italian courts repeatedly refused to validate detention orders following the CJEU’s Grand Chamber ruling in CV,Footnote 145 which held that partial designation of a country as safe, retaining exemptions for specific categories of persons, is incompatible with the Asylum Procedures Directive’s safe country of origin concept.Footnote 146 The Italian executive responded by adopting Decree–Law no. 158/2024 to redefine the safe country list and by transferring jurisdiction to validate detention to the courts of appeal.Footnote 147 Following the Grand Chamber’s judgment in Alace and Canpelli in August 2025,Footnote 148 which confirmed that safety designations must be amenable to full and effective judicial review including ex nunc assessment, the two Albanian centers were re–purposed for pre–removal detention, suspending their use as a purely asylum processing scheme, unless an asylum claim was lodged during the pre–removal detention.Footnote 149 Following this, numerous preliminary references have been sent by Italian courts on the conformity of extraterritorial pre–removal detention with the Return Directive,Footnote 150 which confirm that domestic judiciaries have assumed the role of constitutional gatekeepers, the very dynamic that the Return Regulation risks reproducing at Union scale. The compatibility of extraterritorial detention with EU law is at the moment of writing pending before the Court of Justice in Sedrata, and in Comeri and Sidilli. Furthermore, two Advocates General have already indicated that, while the Regulation’s return centers are legally distinct from the Italy–Albania Protocol, both converge on the same constitutional constraints such as effective judicial scrutiny, the rights of the defense, the prohibition of direct and indirect refoulement, and detention as a last resort tied to a realistic prospect of removal.Footnote 151
The implications for return hubs follow directly from the Court of Justice’s established jurisprudence on non–refoulement and effective remedies. Transfer to a hub state must be preceded by a return decision designating that state as the country of return within the meaning of Article 3, point (3)(g) of the Regulation, thereby triggering the full panoply of safeguards established in Gnandi, FMS and Others, Ararat, and X: a full and individualized ex ante non–refoulement assessment including indirect refoulement; formal notification of the designated country of return; and access to an effective remedy with automatic suspensive effect where an arguable risk under Article 3 ECHR or Article 19(2) CFR is raised.Footnote 152 As the Court made clear in Ararat and X, compliance with non–refoulement must be assessed before enforcement and responsibility attaches at both the point of decision–making, and at the moment of physical transfer. Diplomatic assurances, contractual monitoring clauses, or post–transfer accountability mechanisms cannot substitute for a negative ex ante risk assessment where evidence points to systemic or foreseeable risks of non–refoulement violations in the receiving state.Footnote 153
The key constitutional question is therefore not whether the Union may cooperate with third countries in return matters, but whether the Regulation’s chosen architecture can be implemented compatibly with five essential rule of law requirements: legal form of the agreement; legality and clarity of the return decision; full and individualized ex ante non–refoulement assessment including indirect refoulement; effective judicial protection with automatic suspensive effect when non-refoulement is at issue; legal responsibility that cannot be contractually displaced to a third country, given that the CFR applies in the enforcement of return hubs agreements.
As regards the form of the instrument itself, return hub arrangements must be concluded as formal international agreements pursuant to Article 218 TFEU—when adopted at the EU levelFootnote 154 —because only that procedural framework ensures Council authorization, Parliamentary involvement, and ex ante judicial review. Informal arrangements, whether they are concluded at the EU or national level, whatever their practical convenience, cannot satisfy this requirement, and, critically, their informality does not shield Member State responsibility before the ECtHR, which attaches to the exercise of jurisdiction, not to the legal form of the enabling instrument.Footnote 155
Moreover, the Council’s position codified in the final version of the Regulation, tolerating partial protection deficiencies subject to abstract guarantees, allowing offshore detention, and preserving broad national discretion on suspensive effect, raises serious doubts regarding rule of law conformity. Without precise and enforceable standards on the content of required guarantees, the independence and powers of monitoring mechanisms, and access to counsel and courts from third–country facilities, the return hub model risks amplifying at Union scale the accountability gaps revealed by the Italy–Albania Protocol, placing Member States in a structurally precarious position under both EU law and the ECHR.Footnote 156 This is the extraterritorial logic of exceptionalized pluralism. The further return enforcement is geographically displaced from the constitutional order that governs it, the more the human rights safeguards become, in practice, inaccessible, and the more courts are required to intervene to fill gaps left in the legislative design.
G. Conclusion
The 2024–2026 reform of the EU return acquis crystallizes a constitutional dilemma that two decades of judicial interaction had sought progressively to resolve: how to reconcile the objective of return enforcement efficiency with basic fundamental rights guarantees, in particular, the prohibition of refoulement under Article 19(2) CFR and ill-treatment under Article 4 CFR, the right to liberty under Article 6 CFR, and the right to effective judicial protection under Article 47 CFR. This Article argued that the legislative response is constitutionally inadequate across all four dimensions examined, and that those inadequacies are not incidental design flaws but expressions of a single structural logic: exceptionalized pluralism.
Exceptionalized pluralism, as developed in this Article, describes a regulatory framework in which multiple derogatory procedural tracks, each individually authorized as a response to specific circumstances, collectively displace the common return procedure and its Charter–based safeguards, transforming what was designed as exception into the operative norm of return governance. The concept builds on existing scholarly frameworks that have characterized the direction of EU migration governance as “concerted dis–integration,” “organised hypocrisy,” “de–legalisation,” “crisification,” and “normalisation of disintegration.”Footnote 157 It names a specifically EU return–law mechanism those frameworks have not assessed and which does not fit squarely under any of these theories: the codification of multiple derogations in various directly applicable regulations, which relegates the common return procedure to an exception. The danger is not the single exception but the systemic architecture. When each track qualifies as a formally authorized derogation, the sum of the exceptions becomes the operative norm, and the common return procedure, with its constitutionalized safeguards, is relegated to residual status.
The reform architecture examined in this Article comprises five interlocking instruments adopted in successive stages: the 2024 Asylum Procedures Regulation and Return Border Procedure Regulation, now in force as part of the Pact on Migration and Asylum; the revised Schengen Borders Code, equally in force; and the 2026 Return Regulation, now adopted in its final form.Footnote 158 Each instrument addresses a distinct dimension of return, yet collectively they produce, through their cumulative operation, the architecture of exceptionalized pluralism analyzed here.
The Article has mapped exceptionalized pluralism across its principal dimensions. The merger of asylum and return procedures compresses the procedural space within which non–refoulement obligations falling outside the scope of international protection must be independently assessed, risking their subsumption within asylum reasoning rather than autonomous verification. The normalization of intra–EU transfers at internal borders, through the unreformed Article 6(3) of the Return Directive and the new Article 23a of the revised Schengen Borders Code, codifies enforcement practices that courtsFootnote 159 have consistently found to operate below the constitutional baseline, while preserving the informality and speed that make Charter scrutiny structurally difficult. The mandatory Return Border Procedure embeds containment as a procedural precondition for accelerated return, reconfiguring detention from a last resort into the operational condition of a high–speed track and progressively marginalizing voluntary departure as a genuine alternative. The expansion of pre–removal detention in the Regulation recalibrates the constitutional relationship between liberty and executive power, such that what Article 6 CFR and Kadzoev standards require as exceptional becomes, through cumulative legislative design, the operative baseline. Finally, return externalization through hub–based arrangements projects exceptionalized pluralism extraterritorially. Geographical displacement of enforcement amplifies, rather than reduces, the accountability gaps that Charter applicability to Union–law implementation is designed to prevent, while the Council’s tolerance of partial protection deficiencies subject to abstract guarantees sits in direct tension with the Court’s requirements in Ararat and X that non–refoulement obligations attach ex ante and cannot be contractually displaced.
Across all these dimensions, the constitutional risk does not lie in an absence of formal protection. Non–refoulement remains enshrined in the Regulation, the right to liberty is formally acknowledged, and judicial review is provided for across all return–related measures. Yet Article 28 expressly declines to provide automatic suspensive effect, permitting only discretionary suspension on request, with none for onward appeals. This is precisely the guarantee that, as Gnandi and M.K. and Others v. Poland confirm, effectiveness requires wherever an arguable Article 3 ECHR or Article 19(2) CFR risk is raised. Where accelerated procedures, limited suspensive effect, expanded detention grounds, mutual recognition mechanisms lacking binding fundamental rights suspension clauses, and rule–of–law–deficient externalization multiply differentiated tracks, they collectively compress the procedural space within which residual protection needs can be identified and assessed. That compression is not incidental: it is the structural outcome of a thin legislative design in which individual enforcement decisions are taken too quickly, too remotely, and with too little individualized scrutiny to allow their constitutional constraints to operate in practice. The result is that the absolute prohibition of non–refoulement, formally preserved by each of the return–related instruments, becomes practically vulnerable not through explicit derogation, but through the architecture of speed, distance, procedural thinness, and diffuse accountability that exceptionalized pluralism systematically produces. This concern finds direct support in the Opinion of Advocate General Emiliou in Sedrata. Even where no incompatibility is apparent in the EU legislation as such or between the EU and national legislation, the Advocate General cautioned that difficulties may still emerge in practice because the compatibility of national legislation with the Return Directive “may ultimately depend on the manner in which it is applied by the competent national authorities,” such that any assessment “confined to the letter of the law would … be incomplete.” That caution captures another issue: the constitutional vulnerability of the reformed acquis lies also on in its implementation.Footnote 160
The past decade has demonstrated that judicial correction has been indispensable in constitutionalizing EU return governance. Through more than fifty preliminary rulings, the CJEU embedded proportionality, effective remedies, and non–refoulement into an enforcement–oriented framework the legislature had not initially designed to accommodate them. Yet reliance on ex post judicial intervention cannot substitute for coherent legislative design. Where rights protection depends primarily on litigation, both legal certainty and equal treatment before the law are systematically weakened. Courts operating under conditions of mounting institutional pressureFootnote 161 cannot indefinitely bear the constitutional burden that the legislature declines to assume. That these pressures are directed specifically at courts adjudicating migration and asylum cases is not incidental. It reflects a deliberate attempt to insulate the political choices embedded in exceptionalized pluralism from judicial scrutiny.
The legitimacy of the emerging European return system will not depend on its speed or operational uniformity. It will depend on whether enforcement harmonization is matched by explicit codification of Charter–based safeguards: automatic suspensive effect where non–refoulement is reasonably engaged; a clear and non–delegable obligation on enforcing Member States to conduct independent ex ante non–refoulement assessment; formal and rule–of–law–compliant return hubs agreements; and detention rules that genuinely reflect the CJEU’s settled requirement that deprivation of liberty remains exceptional and proportionate. Without that codification and implementation in practice, exceptionalized pluralism will continue to operate as a self–reinforcing regulatory logic, expanding its derogatory tracks, relegating the common procedure to residual status, and leaving courts the difficult task of acting as gatekeepers of rule of law safeguards.
Acknowledgments
The author would like to thank the ACCESS team members and Professor Marco Borraccetti for helpful comments on earlier versions of this Article.
Competing Interests
The author declares none.
Funding Statement
This work was supported by the European Research Council 2022 Starting Grant Gatekeepers to International Refugee Law—The Role of Courts in Shaping Access to Asylum [101078683].
Open access
